A meta-skill designed to intelligently select, coordinate, and synthesize multiple installed skills to handle complex, multi-domain requests.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install use-skills
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install use-skills using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Use Skills?
The Use Skills capability acts as an orchestrator for your AI agent workspace. Instead of relying on a single domain of expertise, this skill analyzes complex user requests—such as those requiring simultaneous planning, coding, and testing—and selects a tailored working set of Openclaw Skills to provide a unified response.
By implementing a structured selection process and categorized matching (primary vs. support), it ensures that the agent provides high-quality outputs without the noise of irrelevant guidance. It bridges the gap between specialized tools, allowing for a coherent and holistic approach to software development and technical writing tasks.
Use Skills Use Cases
- Orchestrating complex requests that span planning, implementation, and testing.
- Improving documentation structure while simultaneously updating code snippets.
- Performing deep reviews that combine security, performance, and style guidelines.
- Streamlining prompt engineering by layering enhancement skills over domain-specific tasks.
- Managing multi-skill workflows when the user invokes the $use-skills command.
How Use Skills Works
- The skill triggers upon the $use-skills command or when a request clearly benefits from multiple Openclaw Skills.
- It initiates a 'Mode Choice Hard Gate' to let the user select the level of skill involvement (All related, Recommended, or Restricted).
- It reviews the list of visible installed skills and compares them against the user's specific request.
- Skills are labeled as primary (shaping the result) or support (adding quality/edge-case coverage).
- It reads only the relevant portions of the selected skills to maintain focus.
- The skill synthesizes the disparate guidance into one unified answer, plan, or patch.
Use Skills Setup
To enable advanced orchestration, ensure the skill is installed in your agent directory. You can trigger the selection menu manually using:
# Invoke the skill coordination menu
$use-skills
No additional complex configuration is required; the skill automatically detects other visible Openclaw Skills within the session context.
Use Skills Data Schema & Taxonomy
The skill utilizes a structured selection and reporting metadata schema to keep the user informed of the active working set:
| Attribute | Description |
|---|---|
| Mode | The user-selected depth: All related, Recommended, or Restricted. |
| Working Set | A list of bare skill names (e.g., use-skills, brainstorming) currently active. |
| Match Label | Taxonomy used internally to rank skills: primary, support, or skip. |
| Purpose | A brief summary of why the specific set was chosen for the task. |
Use Skills Advanced Features
- Hard Gate logic that forces a mode selection before any tool execution to save tokens and ensure accuracy.
- Conflict handling resolution that prioritizes narrow domain skills over general ones when guidance overlaps.
- Dynamic candidate list generation based on visible session metadata and user-pasted skill blocks.
- Support for aggressive inclusion of auxiliary skills like enhance-prompt for high-fidelity output refinement.
- Adaptive reuse of previously selected modes for recurring tasks to minimize interaction friction.
SKILL.md
Loading
Related Openclaw Skills
Resolution Lattice Trader
An advanced trading agent that exploits mathematical inconsistencies and temporal logic errors across related Polymarket contracts.
AINative Next.js SDK
A server-side SDK for Next.js that simplifies adding streaming AI chat and secure authentication to full-stack applications.
TOSR Test Skill
A specialized integration utility designed to validate the publishing and management lifecycle of AI agent tools.
NeoBoost
NeoBoost is a comprehensive brand intelligence tool for monitoring visibility, health, and AI sentiment across Generative Engine Optimization (GEO) landscapes.
Market MCP
A specialized MCP server providing comprehensive access to China-market financial data and Tushare-oriented datasets for AI agents.
Interviewer Claw
Interviewer Claw is a rigorous, structured AI agent designed to stress-test plans, designs, and ideas by relentlessly probing assumptions until every ambiguity is resolved.
Featured*
